JP2016118547A - Diversification of lenslet, beamwalk, and tilt for anisoplanatic image formation in large diameter telescope - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a system and method in which a standard image reconstruction technology and evaluation of a three-dimensional atmospheric turbulence in tomography are combined in order to enable high-quality anisoplanatic image formation of a distant object from a distant place through atmosphere in a field of view extended by using a large diameter telescope.SOLUTION: Atmospheric tomography and image reconstruction of a forward model, and technology of controlling chip, tile, beamwalk, and focal point are combined in order to generate high-quality image reconstruction of a quick-moving large object when observed with a large diameter telescope in extension of atmospheric turbulence irrespective of its magnitude.SELECTED DRAWING: Figure 3

Description

  The present disclosure relates generally to systems and methods for imaging objects from long distances through the atmosphere, and more specifically to air-to-air and air-to-air monitoring. Specifically, the present disclosure relates to a system and method for compensating for the effects of atmospheric turbulence on images of distant objects (including stars, orbiting objects, missiles and airplanes).

  When light from a star or other celestial body enters the Earth's atmosphere, the image is distorted or moved in various ways by atmospheric turbulence (for example, caused by the interaction of different temperature layers and different wind speeds) There can be. These distortions cause blurring in images generated by any telescope with an aperture larger than 10 cm. Since blurring changes at high speed, higher spatial frequencies are eliminated in images that have been exposed for a long time. One way to address this problem is to correct the wavefront using real time adaptive optics (AO). One complementary method is a speckle imaging technique that reconstructs an image with multiple short exposures.

  The AO system includes a wavefront sensor (WFS) that receives a portion of astronomical light, a deformable mirror located on the optical path, and a computer that receives input from a detector and controls the surface of the deformable mirror. . Alternatively, the deformable mirror can be substituted by other types of devices that provide the ability to adjust the phase of the optical beam as a function of spatial position through the beam and as a function of time. Another example is a liquid crystal spatial light modulator. AO systems rely on wavefront sensors to measure wavefront perturbations caused by turbulence. These measurements can be used to improve the performance of the imaging system by reducing the effects of wavefront distortion. The aim is to correct the deformation of the incident wavefront by quickly deforming the mirror to compensate for the distortion.

  The simplest form of adaptive optics is chip-tilt correction. Chip-tilt correction corresponds to wavefront correction in two dimensions (equivalent to correction of image position offset). Rapid optical tilt in both the X and Y directions is referred to as jitter. Jitter can be generated by rapidly changing three-dimensional (3-D) refraction in an aerodynamic flow field. Jitter can be compensated for in AO systems by using a planar steering mirror attached to a dynamic biaxial mount that allows changes in the X and Y angles of a small, rapid, computer controlled mirror. By reflecting light from a planar steering mirror driven by a computer, the image or laser beam can be stabilized. Image blur due to motion and far-field laser beam jitter is reduced if not eliminated.

  Compensating optics is essentially a coplanar technology, i.e., the effect of turbulence is considered to be constant within an angle called the "isoplanatic angle". In fact, wavefront distortion due to turbulence varies spatially under non-coplanar conditions, while isoplanatic angles vary with atmospheric conditions. Compensating optical correction of extended regions consisting of multiple coplanar patches (ie, optical paths with temporal coherence) requires a significant increase in complexity in AO systems.

  It would be advantageous to provide an astronomical instrument that can perform high-quality non-coplanar imaging in the field of view extended with a large aperture without the increased complexity of the AO system. .

  The approach disclosed herein enables standard image reconstruction techniques and 3D atmospheric turbulence in order to enable high quality non-coplanar imaging in a field of view extended using a large aperture. This is a combination of tomographic evaluation. More specifically, the systems and methods described in detail below provide high-quality image reconstruction of large objects that move quickly when viewed at a diameter larger than about 10 cm in a turbulent dilation, regardless of strength. To create the composition, it combines concepts such as atmospheric tomography, forward model image reconstruction, and diversification techniques.

  Using the atmospheric phase tomography used in adaptive optics, the light collected by the wavefront sensor across multiple frames samples multiple optical paths through the atmosphere, enabling evaluation of atmospheric turbulence volume Will. As long as the forward model is correctly formulated, the same may be true for image reconstruction using a forward model that post-processes the measured data. The measured data depends as a function on the underlying parameters such as the profile of the primitive object. Forward model technology is an algorithm that evaluates the parameters underlying them by matching measurement results with predictions from forward models.

The method proposed here is for image formation of objects at long distances (including objects in the Earth's orbit), and can be easily implemented. It uses the diversity of tilt / beam walks to produce changes that allow for a very detailed phase assessment over time. This approach allows image reconstruction much larger than the “isoplanatic patch angle”. This is because this technique evaluates the entire volume of aberration in the atmosphere between the object and the receiver. This approach utilizes active beam tilt, beam walk and focus control to obtain diversity on demand rather than relying on turbulent layer motion. Using a light receiving sensor (eg, a telescope) constructed in accordance with the teachings of this specification, it may be possible to image an angular region as large as 100 microradians and an aperture of about 100 and a freed length r ₀ .

  The implementation of the non-coplanar imaging system and method disclosed in detail below involves the operation of two fast steering mirrors located at the front of the beam splitter and the lenslet located at the rear of the beam splitter. Array and / or imaging lens actuation may be utilized. The lenslet array is placed in front of the WFS camera and the imaging lens is placed in front of the conventional focal plane array. It is desirable that the image forming camera and the WFS camera sample each image at a fine point density at least as high as that of the Nyquist rate. From an angle perspective, this is λ / 2D. Here, λ is the average wavelength of the operation and image forming camera, and D is the aperture diameter of the optical pupil of the image forming camera. For WFS cameras, D is the diameter of the sub-aperture. However, this is not essential. The system may be able to use cameras with a sampling coarser than Nyquist, although performance may be slightly degraded. In some embodiments, these functions are combined with object motion to allow multi-pass sampling of the atmospheric refractive index volume.

  One aspect of the present subject matter disclosed in detail below is a non-coplanar image forming method of an object from a long distance via the atmosphere, and includes the following steps. (A) accepting light from a distant object at the pupil, (b) forming a continuous focal plane image in the imager camera, and forming a continuous wavefront sensor image in the camera of the wavefront sensor. Aggregating the respective quantities of received light using the imaging technology towards the lenslet array of the imager camera and wavefront sensor; (c) evaluating the atmospheric phase volume and the point spread function of each wavefront sensor image; Using forward model techniques to update jointly, (d) reconstructing images of distant objects based on the updated point spread function and atmospheric phase volume evaluation, and (e) reconstruction The steps from (a) to (d) are repeated repeatedly until the atmospheric phase aberration is minimized in the resulting image, and the diversification technique is controlled by the computer system. The process (c) and (d) are performed by a computer system. According to some embodiments, the diversification technique includes changing the image of the pupil of the wavefront sensor camera.

  Another aspect of the disclosed subject matter is a non-coplanar image forming method for an object from a long distance via the atmosphere, wherein (a) light from a distant object is passed through an aperture of a light receiving sensor. Receiving, (b) separating the received light into a first beam directed to the imager camera and a second beam directed to the lenslet array of the wavefront sensor, (c) the camera of the wavefront sensor Changing the image of the pupil on the focal plane array of (d) evaluating three-dimensional volume atmospheric phase aberrations within the field of view of the multiple sub-aperture light receiving sensor for each image output by the wavefront sensor camera; and (E) Reconstructing a non-coplanar image of a distant object using the estimated atmospheric phase aberration, wherein process (c) is controlled by a computer system and processes (d) and (e) Comprising, performed by a computer system. This method may further include one or more of the following operations. (I) measuring various partially overlapping light guides in a plurality of directions, wherein process (d) evaluates the sum of a plurality of independent phases in a turbulent layer and the atmospheric phase aberration of a three-dimensional volume Measuring, including solving using linear techniques to: (ii) in order to cause partial overlap of sub-aperture images in successive images acquired by a wavefront sensor camera Applying diversity, wherein the process (d) includes stitching the sub-aperture wavefronts using a forward model technique; (iii) aligning the lenslet array along the optical axis of the wavefront sensor. Applying a diversity of focus by adjusting the focus of the WFS like moving, and the operation (e) is on a non-coplanar surface of each sub-aperture image Including, applying, and (iv) applying focus diversity by adjusting the focus of the imaging system such as moving the imaging lens or camera of the imaging system along the optical axis of the imager. Applying, wherein operation (d) includes an evaluation of the phase of a distant object extending across a plurality of coplanar patches, and (v) decomposing the apertures into regularly arranged sub-apertures. A large number of sub-aperture images are formed from the image output by the wavefront sensor camera.

  Yet another aspect of the present subject matter disclosed in detail below is an imaging system from a long distance via the atmosphere, including: A light receiving sensor having an opening, a first high-speed steering mirror that is disposed on the pupil plane and receives light that has passed through the opening, is disposed at a distance from the pupil plane, and reflects light from the first high-speed steering mirror. A second high-speed steering mirror for receiving, a beam splitter arranged to receive the reflected light from the second high-speed steering mirror, and separating the incoming light into first and second beams; receiving the first light beam A lenslet array arranged to receive, a wavefront sensor camera arranged to receive light passing through the lenslet array, an imager camera arranged to receive the second light beam, and the following operations: A computer system that is programmed to run. (A) control of the operation of the first and second high-speed steering mirrors and / or lenslet array to cause a change in the image of the pupil of the wavefront sensor camera, (b) the image captured by the imager camera or the wavefront sensor camera (C) Update of the evaluation of the three-dimensional atmospheric phase volume based on the sub-aperture image captured by (c) Update of the evaluation of the point spread function based on the image captured by the imager camera or the sub-aperture image captured by the wavefront sensor camera, and ( d) Non-coplanar image reconstruction of the object using updated point spread function and atmospheric phase volume estimation. The computer system controls the operation of the lenslet array in parallel and perpendicular to the optical axis of the wavefront sensor camera. The computer system is adapted to control the operation of the first and second fast steering mirrors to produce a change in the sub-aperture images of successive samples captured by the wavefront sensor camera to provide beam walk diversity. , And operation (c) can be further programmed to include stitching the phase data using forward model techniques. Operations (b) through (d) are performed iteratively using forward model techniques until atmospheric phase aberrations in the reconstructed image are minimized. The forward model may include a penalty function for atmospheric turbulence that is inconsistent with expected turbulence statistics and location and time correlation coefficients.

  Other aspects relating to non-coplanar imaging systems and methods via the atmosphere of distant objects are disclosed and claimed below.

FIG. 2 is a diagram representing components of a non-coplanar imaging system for an object from a long distance via the atmosphere, according to one embodiment. The object is shown in three different positions over time. The area within the circle shows a partial overlap of the multiple light paths that the telescope receives from the object at various times, thereby allowing an evaluation of the 3-D atmospheric turbulence profile. FIG. 2 is a diagram representing components of a non-coplanar imaging system for an object from a distance via the atmosphere, according to an alternative embodiment. The area within the circle shows a partial overlap of multiple light paths that are simultaneously received from the object, thereby allowing evaluation of the 3-D atmospheric turbulence profile. 5 is a flowchart illustrating steps of a non-coplanar image formation process according to one embodiment, including operations performed by a processor that receives focal plane and WFS images and controls the diversity applied to the images. is there. FIG. 6 is another flowchart showing in more detail the operations performed by the post processor. In the following, reference is made to the figures, wherein like elements in different figures are given the same reference numerals.

  Various embodiments of systems and methods for non-coplanar imaging of distant objects, including Earth-orbiting objects, with compensation for distortions caused by atmospheric turbulence will be illustrated in detail below. Let's go. These non-coplanar imaging systems are used to create high-quality image reconstructions of large objects that move quickly when observed at large apertures in turbulent expansions, both strong and weak. It combines photography, forward model image reconstruction, and diversification techniques.

  Several components of a first exemplary embodiment of a non-coplanar imaging system are schematically depicted in FIG. 1 and show objects moving in a stationary atmosphere and telescope field of view (FOV). . (An alternative description of sensor geometry is to convert the atmosphere and objects in place in the FOV of a swiveling telescope. For example, when tracking a low earth orbit (LEO) satellite) Observation with a telescope The object 2 to be played is shown at three different positions in Time1, Time2, and Time3, respectively. The spotted band in FIG. 1 also represents a “pipe” of light of each of the varying phases that originates from object 2 in Time 1, Time 2, and Time 3. [As used herein, the term “pipe” means a volume of pipe-shaped volume along the viewing direction of an object. Light from the object can have phase aberrations due to atmospheric turbulence in the pipe. In the example shown in FIG. 1, the phase pipe partially overlaps (at different times) in two regions 28 and 30, each represented by an ellipse. Light in the field of view is captured by the telescope and used to form an image of the object 2. However, since there is nothing to compensate for the effects of interfering turbulence 4 (shown by the wavy lines in FIG. 1), the image may be blurred or otherwise distorted.

  A system and method for compensating for phase aberrations caused by turbulence in the light from the original object will now be described with reference to FIG. FIG. 1 includes a block diagram showing components of a non-coplanar image forming system. The telescope includes a primary mirror 40 that defines a light receiving aperture or pupil 6, and a secondary mirror 42. Light incident on the main mirror 40 is reflected toward the sub mirror 42, and the sub mirror then reflects the collision light toward the folded plane mirror 8. The light from the secondary mirror 42 is then reflected by the folded plane mirror 8 toward the first relay lens system 10. When light is emitted from the first relay lens system 10, it collides with the first high-speed steering mirror 12 to form an image of a pupil that coincides with the first high-speed steering mirror 12. The first high speed steering mirror 12 reflects this incident light toward the second high speed steering mirror 14. The second high-speed steering mirror 14 is located at a distance Δz from the pupil image. The pupil image on the second high-speed steering mirror 14 is blurred.

  In accordance with the embodiment illustrated in FIG. 1, the second fast steering mirror 14 is oriented so that the light reflected therefrom collides with a beam splitter 16 (eg, a dichroic or photometric beam splitter). The light is split into two by the beam splitter 16. Part of the light is reflected towards the imaging lens 15 of the imager camera 18 and the other part of the light is transmitted towards the second relay lens system 20. The imager camera 18 further includes a focal plane array 17 that receives light that has passed through the imaging lens 15. Focal plane array 17 converts the focal plane image into an electrical image signal that can be read, processed and / or stored by processor 26. Preferably, but not essential (as described above), the focal plane is sampled at the Nyquist frequency.

  With continued reference to FIG. 1, light that has passed through the second relay lens system 20 impinges on a lenslet array 22 of Shack-Hartmann or other types of wavefront sensors. The wavefront sensor further includes a WFS camera 24. The lenslet array 22 comprises a set of lenslets (not shown) arranged on the pupil plane. Each lenslet is an individual lenslet having the same focal length, and the lenslet group is arranged linearly or regularly on the pupil plane. The lenslet helps to decompose the pupil 6 to form a sub-aperture represented by the parallel vertical lines of FIG. Each lenslet is focused on a focal plane array (for example, a CCD array) of the WFS camera 24. The focal plane array (not shown in FIG. 1) of the WFS camera 24 is an electrical image signal from which the processor 26 can read, process and / or store multiple sub-aperture images corresponding to multiple lenslets. Convert to By using a Shack-Hartmann wavefront sensor that measures tilt, it is possible to measure the phase and amplitude of the bright field.

  In accordance with one embodiment of wavefront reconstruction, processor 26 is programmed to calculate the local tilt of the wavefront through each lenslet from the position of each focus on the focal plane array of WFS camera 24. Can. Any phase aberration can be approximated by a separate set of tilts. By sampling the light passing through the lenslet, the tilt can be measured and any wavefront phase aberration can be approximated by the processor 26.

  During operation of the telescope, the high speed steering mirrors 12 and 14, the lenslet array 22, and (optionally) the imaging lens 15 of the imager camera 18 are activated. The X and Y angles of the first fast steering mirror 12 can be controlled by the processor 26 (or another computer or processor) to achieve tilt diversity. The angles of the first and second fast steering mirrors 12 and 14 can be controlled by the processor 26 to provide beam walk diversity. (As used herein, the term “beam walk” means a change in the image of the pupil on the wavefront sensor.) In addition, the lenslet array 22 has a focal point as shown by the double arrows in FIG. In order to achieve this (ie lenslet) diversity, it can be moved back and forth along the optical axis of the WFS camera 24. The lenslet array 22 is also movable back and forth along an axis orthogonal to the optical axis of the WFS camera 24 (not shown in FIG. 1) to provide beam walk diversity. Optionally, the imaging lens 15 is also movable back and forth along the optical axis of the imager camera 18 to achieve focus diversity. The processor is programmed to control the fast steering mirror, lenslet array and imaging lens to adjust tilt, focus, and beam walk during the process of iterative image acquisition and image reconstruction. The

  In the system shown in FIG. 1, the diversity of tilt, focus, and beam walk coupled with object motion allows multi-pass sampling of the refractive index volume of the atmosphere. This multi-pass sampling allows the processor 26 to remove distortions caused by turbulence from the reconstructed object image using tomographic evaluation of three-dimensional atmospheric turbulence. This operation performed by processor 26 will be described in more detail later in connection with FIG.

  The movement of the original object 2 has the result that the angle of the atmospheric phase sample varies as shown in FIG. The WFS camera 24 can sample the light from the object 2 by sampling the atmosphere in various directions. The phase is measured along the “pipe” represented by the spotted band in FIG. Where the pipes intersect each other is indicated by an ellipse in FIG. Light emitted at different times from the same object point encounters different volume turbulences that induce different phase aberrations and produce point spread functions with different shapes. In regions 28 and 30, each turbulent layer can be solved by using a linear technique to evaluate the 3-D atmospheric phase volume by measuring various overlapping pipes in multiple directions. A sum of multiple independent phases is created.

  By adding the diversity of beam walks (given by the lateral movement of the second fast steering mirror 14 and / or lenslet array 22), sub-aperture partial overlaps can be re-established. It is possible to “stitch” the configuration. As shown in FIG. 1, the high-speed steering mirror 14 is not located in the pupil image, and the pupil image is adjusted by adjusting the angle of the mirror (and compensating the angle by the high-speed steering mirror 12). Can optionally move onto the WFS camera 24. This results in partial overlap of successive WFS sample sub-apertures and allows the processor 26 to stitch the phase information using well-known forward model image and phase reconstruction techniques. Beamwalk also provides the ability to measure wavefront tilt in a spatially denser set of position samples. For example, transforming the beam across the lenslet array with a half-width lenslet provides a new set of measurement results at an intermediate sample point for the previous position.

  Adding tilt diversity (provided by the first fast steering mirror 12) allows fine sampling of the focal plane to enable super-resolution. Adjusting only the fast steering mirror 12 in the pupil image results in a change in the focal plane image on the WFS camera 24. By changing the tilt, the alignment of the beam of each lenslet to the detector array is adjusted. Making measurements on different sets of tilts provides a collectively dense set of samples compared to a single measurement. This allows fine sampling of the sub-aperture image, resulting in super-resolution of the image within the sub-aperture. This can be achieved by using any one of a plurality of techniques. A high-resolution image of the focused beamlet of each lenslet provides not only the tip and tilt recovery of the beam's wavefront profile for each lenslet, but also high focus, astigmatism, coma, and more complex profiles. Supports the recovery of the next structure.

  Adding the degree of freedom to provide focus diversification includes moving the lenslet array 22 along the optical axis of the WFS camera 24. This type of motion allows a more direct non-coplanar image reconstruction at each sub-aperture. This focus diversification technique can also be applied to the focal plane array of the imager 18 to evaluate the phase of an object extending across a plurality of coplanar patches.

  This focal diversification alternative, as shown in FIG. 2, provides a second path to large aperture non-coplanar imaging when used with a Shack-Hartmann sensor (ie, lenslet array 22 and WFS camera 24). provide. In order to avoid the complexity of the figure, in FIG. 2, the phase pipe is represented by a two-dot chain line instead of a spotted band. Light emitted from different object points at the same time encounters different volume turbulences that induce different phase aberrations and produce point spread functions with different shapes. In the example shown in FIG. 2, several phase pipes partially overlap in three regions 32, 34, and 36, indicated by respective ellipses. Different point spread functions obtained from different points of the object can be used to reconstruct the wavefront from these different object points. These wavefronts can unwrap from the modulo 2π phase to generate phase aberrations. The phase then matches the integrated phase along the phase pipe illustrated in FIG. These integrated phases, along with the intersections illustrated by regions 32, 34, and 36, can be utilized to determine the 3-D volume of the phase, as in the embodiment described in FIG. . Light in the WFS field of view is captured by the telescope and used to generate a plurality of sub-aperture images corresponding to each lenslet that makes up the lenslet array 22. By evaluating the 3-D atmospheric phase volume by measuring various partially overlapping pipes, such as those intersecting in regions 32, 34, and 36, using a point spread function in this alternative approach. A number of independent phase sums through each turbulent layer that can be solved is created. This approach described in FIG. 2 may provide such an image in only two frames of data. This approach is therefore one of the possible paths to non-coplanar imaging of larger objects with shorter waiting times.

  FIG. 3 is a flowchart illustrating steps of a non-coplanar image forming process according to one embodiment. The large-aperture telescope receives light in its field of view (FOV) while pointing at an orbiting object (step 50). The light is caused to form a beam by the aperture (ie, pupil) of the telescope. This beam is then affected by various tilts and changes due to the actuation of the high speed steering mirror (step 52). The operation of the fast steering mirror is controlled by a processor or other computer system. After passing through the high speed steering mirror, the light is split into two (step 54). The light reflected by the beam splitter is directed to the focal plane of the imager camera. The imager camera captures images continuously by converting photons into charges (step 62). The light that has passed through the beam splitter is then passed through a lenslet array, which is actuated to assume various axial and lateral positions to provide on-demand diversity of lenslets and beam walks. (Step 56). The operation of the lenslet array is controlled by a processor or other computer system. The lenslet array also has the effect of breaking the apertures into a regular array of sub-apertures (step 58). Each lenslet focuses light on the focal plane of the WFS camera. The WFS camera captures multiple sub-aperture images by converting photons into charges (step 60). The captured images and sub-aperture images are sent to a processor or other computer system. A processor or other computer system uses the image data to evaluate the phase aberration due to atmospheric turbulence (step 64). Processors and other computer systems use phase estimation to reconstruct a less blurred image of the object (step 66).

  The tilt, beam walk and focus diversity described above eliminates uncontrolled diversity including turbulent layer motion. This type of uncontrolled diversity typically requires the evaluation of extra harmful parameters associated with random changes in collection conditions (eg, wind changes).

  As illustrated in FIG. 4, the processor continuously receives a continuous focal plane image 70. Each focal plane image 70 is corrected for pixel gain and offset to provide a more uniform image (operation 72). This correction scheme adjusts the signal response of all pixels of the image based on two calibration points. One of the calibration points is a zero signal (offset) and the other is typically about 50% saturation (for gain calculation). Pixel gain is how the amount of signal output by the detector changes as a function of the amount of light impinging. The processor then performs pixel thresholding and scene support on the corrected focal plane image (operation 74). Pixel threshold processing is an operation in which a pixel having a value below a certain threshold is treated as noise and set to be equivalent to zero. Scene support includes setting the value of a pixel corresponding to a location that a person may wish to ignore in the scene to zero. The metric of the forward model of the imager is then updated with the new focal plane image pixel values (operation 76).

  With continued reference to FIG. 4, the processor simultaneously receives a continuous wavefront sensor image 80 at the same time. Each wavefront sensor image 80 is corrected for pixel gain and offset to provide a more uniform image (operation 82). Thereafter, the processor performs pixel threshold processing and scene support for the corrected wavefront sensor image (operation 84). The metric of the wavefront sensor forward model is then updated with the pixel values of the new wavefront sensor image (operation 76). Each new wavefront sensor image is decomposed into a number of sub-aperture images.

  A portion surrounded by a broken-line square in FIG. 4 represents execution of the forward model. The processor 26 (see FIGS. 1 and 2) implements a first least squares method that minimizes the sum of the squares of the residuals associated with the forward model, forming the response of the imager 18 and the response of the wavefront sensor 24. It is programmed to run. These forward models can be implemented individually or can be combined into one forward model. The forward model metrics are minimized to produce an updated point spread function estimate (operation 78) and an updated atmospheric phase volume estimate (operation 88). These processes consist of updating the evaluation of the three-dimensional atmospheric phase volume based on the image captured by the imager camera or the sub-aperture image captured by the wavefront sensor camera, the image captured by the imager camera, or the sub-aperture image captured by the wavefront sensor camera. Update of the point spread function evaluation based on, and reconstruction of non-coplanar images of the object using the updated point spread function and atmospheric phase volume evaluation. (As is well known, the point spread function can be calculated directly from the phase, and the recovered phase can be checked against the associated point spread function.) To best fit each data set The purpose is to adjust the parameters of the forward model. The least square method finds an optimal solution when the residual sum of squares (hereinafter referred to as “metric”) is minimized. ("Residual" is defined as the difference between the measured pixel value and the pixel value predicted by the forward model.) This nonlinear problem is solved by iterative refinement. Point spread function updates and atmospheric phase updates are used to generate an updated image after each iteration (operation 90). Prior to the next iteration, adjustments (ie, application of tilt, focus, and beamwalk diversity) are made to the hardware as needed (operation 92) under processor control. After each set of adjustments, additional images are acquired and processed until the nonlinear minimization problem is solved. As a result, the updated image has minimal atmospheric phase aberration.

  Other innovations in the imaging methods disclosed herein include: (a) layers obtained by wind or other turbulent development, including forward model penalty functions for large winds or complex developments And (b) a turbulence penalty function that does not match the expected turbulence statistics, for example, a penalty for turbulence evaluation that does not follow Kolmogorov statistics.

  The image reconstruction technique employed herein uses jointly an image sequence formed by both the conventional image forming path and the Shack-Hartmann (lenslet) path. Known iterative maximization techniques can be used to determine the reconstructed object that best matches the raw pixel level data, both conventional and Shack-Hartmann.

Using a wavefront sensor in combination with a focal plane image to test tilt or test an object viewed through a partial area of the pupil has several benefits, including reduced redundant noise and reduced aberrations. The performance should be good up to D / r ₀ = 100 (assuming about 10 sub-apertures over the entire aperture with D / r ₀ ≈10 and reconstruction within each sub-aperture). The proposed approach will work well under conditions of larger objects moving at high speeds, large apertures, and extended or intense turbulence (corresponding to low elevation angles).

ここでは、和はｋ＝１からＫまでの層のそれぞれの位相の和であり、ｎ（ｉ、ｊ、ｍ）は対応する測定ノイズである。シャックハルトマン波面センサでは、測定されるのは位相のチルトである。

ここでは、δΤ（ｉ、ｊ、ｍ）は測定されたチルトであり、δΤ_Ｌ（ｉ、ｊ、ｋ；ｍ）は評価対象の層のチルトである。ｍ＝１からＭまでの累積は、限られた未知数δΤ_Ｌ（ｉ、ｊ、ｋ；ｍ）を有する多数の等式の集合という結果に帰着するように見える。δΤ_Ｌ（ｉ、ｊ、ｋ；ｍ）においては、指数ｉ及びｊは、暗黙のうちに測定時間ｍの関数であり、したがってグリッド化の問題が現存する。 The basic approach shown in FIG. 1 is to measure the sum of the phase tilts along multiple paths of each sub-aperture. The measurement result δφ (i, j, m) of the phase of the (i, j) -th sub-aperture at time m is expressed by the following formula.

Here, the sum is the sum of the respective phases of the layers from k = 1 to K, and n (i, j, m) is the corresponding measurement noise. In the Shack-Hartmann wavefront sensor, it is the phase tilt that is measured.

Here, δΤ (i, j, m ) is the measured tilt, and δΤ _L (i, j, k; m) is the tilt of the evaluation target layer. The accumulation from m = 1 to M appears to result in a set of multiple equations with a limited unknown δΤ _L (i, j, k; m). In δΤ _L (i, j, k; m), the indices i and j are implicitly a function of the measurement time m, so the problem of gridding exists.

  For objects in Earth's orbit, the dominant movement is attributed to the object and the wind effect is secondary. Thus, the ray path is relatively well known over time, thereby minimizing gridding problems. The intersection of light paths 28 and 30 in FIG. 1 (shown as an ellipse) represents a partially overlapping volume cell. This partially overlapping volume cell allows the resolution of the corresponding individual phases that make up the path integration.

  One of the key additional features of the non-coplanar imaging system disclosed herein is as part of image reconstruction (or image reconstruction) rather than as part of AO wavefront reconstruction. The phase is evaluated (apart from construction). The proposed image reconstruction typically involves integrating light against both the imager camera 18 and the focal plane of the WFS camera 24.

  Additional features of non-coplanar image forming systems include: First, the phase can be evaluated by a forward model technique that allows for larger sub-apertures, such as multiframe blind deconvolution at each sub-aperture. Second, the relative average phase of the sub-apertures is not directly known within a given framework. The average of these unknown phases can be evaluated by changing the image of the pupil of the wavefront sensor and seeing the partial overlap of the phase region with successive measurements. This pupil image change (which is an example of a beam walk) is achieved by using a fast steering mirror 14. The effect of the high speed steering mirror 14 is to move the pupil image by an amount of ΔzΔθ. Here, Δθ is an angular displacement on the high-speed steering mirror (the enlargement effect is ignored for simplification), and Δz is the path length of the high-speed steering mirror 14 from the image plane of the pupil. The effect of the fast steering mirror 14 will also cause a tilt that changes the focal plane image position. This can be canceled by using the high-speed steering mirror 12. The fast steering mirror 12 can also be used alone to apply tilt and thus apply focal plane image changes, but no pupil image changes occur. The benefit of image change is that it allows sampling of the image at an angular interval that is finer than the angular width of the pixel. This supports super-resolution sampling of the entire sub-aperture image or focal plane image. Finally, the focus (ie, lenslet) diversity is applied by moving the lenslet array 22 relative to the focal plane of the WFS camera 24 (as indicated by the double arrows in FIG. 1). Past studies have demonstrated that this diversity of focus supports non-coplanar imaging (image formation in multiple coplanar patches).

In accordance with non-coplanar imaging disclosed herein, the image reconstruction process is enhanced by the movement of two fast steering mirrors and the operation of the image lens and lenslet array. Due to these changes, a forward model evaluation of the total volume of phase aberrations in the atmosphere is possible with only a slight movement of the object. This slight angular motion must satisfy δθ> d _sub / H. Here, d _sub is the width of the sub-aperture when mapped to the pupil of the telescope, and H is the height of the atmosphere. Thus, for example, when d _sub = 30 cm and H = 20 km, the required object motion is only about δθ> 15 microradians in the output space. The object motion required to identify the lowest altitude turbulent layer, typically within 1 km of the telescope, is about 300 microradians. This can be achieved with about 5 seconds of data for stellar object motion and 0.1 seconds for LEO objects. If signal-to-noise ratio (SNR) is an issue, longer history may also be effective. The relatively fast time of 0.1 second for LEO objects indicates a significant smearing of phase information for a typical frame rate of 100 Hz. Thus, a faster frame rate may be appropriate.

  The change provided by the two high-speed steering mirrors allows sub-pixel sampling of the object plane and sub-pixel sampling of the pupil plane. This allows for a finer evaluation of features than would normally be obtained with that type of pixel, or conversely, larger pixels and sub-apertures than would normally be needed to obtain a given resolution. It becomes possible. This benefits SNR.

The non-coplanar image forming system proposed in this specification employs a forward model that uses the tilt δΤ _L (i, j, k; m) of the atmospheric phase (difference in the optical path of the atmosphere) as an independent variable. To do. By using tilt rather than phase, the conditioning of the problem will be improved. This is because the phase has an overall piston that is not sensed, which results in a single set of equations. Tilt is evaluated in the forward model simultaneously with image reconstruction. The use of tilt or phase difference is a natural choice for image reconstruction. This is because the image point spread function is also a function of the phase difference (for each patch on the same plane).

  The 3-D grid used to represent the volume of the atmosphere can be coarse for faster reconstruction. Alternatively, a finer 3-D grid may be used for more accurate reconstruction, but the reconstruction time will result in longer time.

  The system proposed here is expected to enable imaging of angular regions exceeding 100 microradians. At any given time sample, if focus diversification is used, a field of view (FOV) compensation approximately equal to the overall FOV of the imaging system for a scenario looking up in weaker turbulence is achieved. Is expected to be. This FOV is about 100 microradians (this assumes the use of a state-of-the-art camera of about 1000 x 1000, assuming a Nyquist sampling of 100 nanoradians). In the case of stronger turbulence, diversification of focus may not be sufficient by itself. In such cases, the volume turbulence technique employed herein may be helpful. And if strip-like data with a width of 300 microradians is obtained, an image of that size (300 microradians) can be formed in the direction along the trajectory, and the width of the object in the direction across the trajectory. Equal widths may be imageable (although for very large objects, a large FOV sub-aperture may be required).

The proposed method is capable of evaluating high quality images that are much larger than the patch angle on the same plane, with a size of 100 microradians or more and D / r ₀ of 100 or more. Should be. There are also potential benefits to SNR. This is because signals from a plurality of frames are combined to form an image. If the super-resolution function of the forward model is fully utilized to obtain super-resolution for both image and turbulence features, the image can be of exceptionally high quality. Be expected. Rather than relying on the assumption that turbulence is layered and moving at a constant speed, this approach is more robust than previously proposed approaches due to the user-specified variety of image changes and focus. It is expected to be. Rather, this approach can test those assumptions in creating a turbulent volume assessment.

  In addition, the proposed approach will make it possible to provide an improved phase estimate that improves the quality of the laser light beam transmitted via the atmosphere. This type of adaptation would provide performance benefits, for example, for laser or optical communication systems that attempt to establish a link between a terrestrial source terminal and a destination terminal on an Earth orbiting satellite. Improved non-coplanar phase correction also provides improved satellite illumination from the ground or aircraft for active or laser imaging, providing optical power transmission, or laser weapon applications I can do it.

  Finally, the proposed technique says that for imaging an object in low earth orbit, if the object's slew is much faster than the wind speed at that location, the layers moving separately are less relevant. It is a premise and should be particularly useful.

  While non-coplanar imaging systems and methods of distant objects via the atmosphere have been described with reference to various embodiments, those skilled in the art will recognize various methods without departing from the teachings herein. It will be appreciated that variations are possible and that elements may be substituted with equivalents. In addition, for a particular situation, numerous modifications may be possible to adapt concepts and simplifications to the practices disclosed herein. Thus, it is intended that the subject matter claimed is not limited to the disclosed embodiments.

  As used in the claims, the term “computer system” includes a system having at least one computer or processor and a system that may have multiple computers or processors communicating over a network or bus. Should be interpreted widely. The terms “computer” and “processor” used in the previous sentence both refer to a processing device (eg, a central processing unit) and some form of memory (ie, a computer-readable medium) that stores a program that can be read by the processing device. Means a device having These terms also include digital and analog devices that perform signal processing functions that take one or more signals, apply mathematical functions to filter the signals, and generate one or more output signals. Also intended.

  In addition, the claims of the method specified below are such that the steps listed therein are performed in alphabetical order (any alphabetical order in the claims simply refers to the steps listed previously). Should not be construed as having to be performed in the order listed. They should also not be construed as excluding the execution of any part of two or more steps simultaneously or interchangeably.

Claims (14)

A non-coplanar image forming method of an object from a long distance via the atmosphere, the method comprising: (a) receiving light from a distant object at a pupil;
(B) to form a continuous focal plane image in the imager camera and a continuous wavefront sensor image in the camera of the wavefront sensor, the respective amount of the light received using diversification techniques; And aggregating towards the lenslet array of the wavefront sensor,
(C) jointly update the evaluation of the atmospheric phase volume of each wavefront sensor image, and use forward model technology to update the evaluation of the point spread function;
(D) reconstructing an image of the distant object based on the updated point spread function and atmospheric phase volume evaluation; and (e) minimizing atmospheric phase aberrations in the reconstructed image. Including repeating steps (a) to (d) repeatedly,
A method wherein the diversification technique is controlled by a computer system and processes (c) and (d) are performed by the computer system.   The method of claim 1, wherein the diversification technique includes changing an image of the pupil of the wavefront sensor camera.   The method of claim 2, wherein the image of the pupil is changed by moving the lenslet array in a direction perpendicular to the optical axis of the wavefront sensor camera.   The method of claim 2, wherein the image of the pupil is changed by changing the orientation of one or more fast steering mirrors.   The method of claim 1, wherein the diversification technique includes tilt diversification.   The method of claim 1, wherein the diversification technique includes focus diversification.   The image of each wavefront sensor includes a plurality of sub-aperture images, and step (c) includes measuring a total tilt along a plurality of paths and evaluating a tilt of each layer of the atmosphere. The method described.   Step (c) further evaluates the atmospheric turbulence movement in each layer, obtained by wind or other development of the atmospheric turbulence, including penalty functions in forward model techniques for large wind and complex deployments The method of claim 7 comprising:   The method of claim 7, wherein the forward model technique includes an atmospheric turbulence penalty function that is inconsistent with expected turbulence statistics. An image forming system for an object from a long distance via the atmosphere,
A light receiving sensor having an opening,
A first high-speed steering mirror disposed on the pupil plane and receiving light that has passed through the opening;
A second fast steering mirror disposed at a distance from the pupil plane and receiving reflected light from the first fast steering mirror;
A beam splitter arranged to receive the reflected light from the second high-speed steering mirror and separating the incoming light into first and second light beams;
A lenslet array arranged to receive the first light beam;
A wavefront sensor camera arranged to receive light that has passed through the lenslet array;
An imager camera arranged to receive the second light beam, and a computer system,
(A) an operation for controlling the operation of the first and second high-speed steering mirrors and / or the lenslet array for causing a change in the image of the pupil of the wavefront sensor camera;
(B) an operation for updating an evaluation of a three-dimensional atmospheric phase volume based on an image captured by the imager camera or a sub-aperture image captured by the wavefront sensor camera;
(C) an operation for updating an evaluation of a point spread function based on an image captured by the imager camera or a sub-aperture image captured by the wavefront sensor camera; and (d) an evaluation of the updated point spread function and atmospheric phase volume. A computer system programmed to perform an operation to reconstruct a non-coplanar image of the object using.   The system of claim 10, wherein the computer system controls the operation of the lenslet array in parallel and perpendicular to the optical axis of the wavefront sensor camera.   To control the operation of the first and second fast steering mirrors to provide beam walk diversity that causes partial overlap in the sub-aperture images of successive samples captured by the wavefront sensor camera. The system of claim 10, wherein the computer system is further programmed and the operation (c) comprises stitching the phase data using forward model techniques.   11. The system of claim 10, wherein operations (b) through (d) are performed iteratively using forward model techniques until atmospheric phase aberrations in the reconstructed image are minimized.   14. The system of claim 13, wherein the forward model includes an atmospheric turbulence penalty function that is inconsistent with expected turbulence statistics.

Images